From: sashiko-bot@kernel.org
To: "Edward Srouji" <edwards@nvidia.com>
Cc: linux-rdma@vger.kernel.org
Subject: Re: [PATCH rdma-next 04/11] RDMA/mlx5: Select the core device through PD for QP commands
Date: Wed, 18 Nov 2026 14:04:16 +0000 [thread overview]
Message-ID: <sashiko-outbox-221133@kernel.org> (raw)
In-Reply-To: <20261117-mlx5-sd-prep-v1-4-8efd124c7895@nvidia.com>
Thank you for your contribution! Sashiko AI review found no new issues
in this patch, but noticed 2 related pre-existing issues it does not
introduce:
- [Critical] RDMA/mlx5: unhandled error in destroy_raw_packet_qp_rq()
causes use-after-free [1]
- [High] RDMA/mlx5: unchecked to_ib_qp_state() return in query_qp_attr()
[2]
--
commit c8de8977935b4bc94238c72869a744ae61d62546
Author: Akiva Goldberger <agoldberger@nvidia.com>
RDMA/mlx5: Select the core device through PD for QP commands
This commit introduces qp_to_mdev() to resolve the core device from a QP's
associated PD and updates QP, RQ, SQ, and DCT command helpers to obtain the
core device from the PD or caller.
> diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
> index da4e9a52d2f9b..2196e06f593e1 100644
> --- a/drivers/infiniband/hw/mlx5/qp.c
> +++ b/drivers/infiniband/hw/mlx5/qp.c
[ ... ]
> @@ -1526,17 +1525,16 @@ static int create_raw_packet_qp_rq(struct mlx5_ib_dev *dev,
> pas = (__be64 *)MLX5_ADDR_OF(wq, wq, pas);
> mlx5_ib_populate_pas(umem, page_size, pas, 0);
>
> - err = mlx5_core_create_rq_tracked(dev, in, inlen, &rq->base.mqp);
> + err = mlx5_core_create_rq_tracked(pd, in, inlen, &rq->base.mqp);
>
> kvfree(in);
>
> return err;
> }
>
> -static void destroy_raw_packet_qp_rq(struct mlx5_ib_dev *dev,
> - struct mlx5_ib_rq *rq)
> +static void destroy_raw_packet_qp_rq(struct ib_pd *pd, struct mlx5_ib_rq *rq)
> {
> - mlx5_core_destroy_rq_tracked(dev, &rq->base.mqp);
> + mlx5_core_destroy_rq_tracked(pd, &rq->base.mqp);
> }
[Severity: Critical]
This problem wasn't introduced by this patch, but can ignoring the return
value of mlx5_core_destroy_rq_tracked() in destroy_raw_packet_qp_rq()
[1] cause a use-after-free?
When the firmware DESTROY_RQ command fails, mlx5_core_destroy_rq_tracked()
re-enables the resource and leaves &rq->base.mqp in dev->qp_table.tree:
drivers/infiniband/hw/mlx5/qpc.c:mlx5_core_destroy_rq_tracked() {
...
modify_resource_common_state(dev, rq, true);
ret = destroy_rq_tracked(pd_to_mdev(pd), rq->qpn, rq->uid);
if (ret) {
modify_resource_common_state(dev, rq, false);
return ret;
}
destroy_resource_common(dev, rq);
return 0;
}
Because destroy_raw_packet_qp_rq() drops that error, raw packet QP teardown
completes and frees the enclosing mlx5_ib_qp. If an asynchronous firmware
event later arrives for the undestroyed RQ, rsc_event_notifier() looks up
the dangling pointer in the radix tree and calls qp->event() on freed
memory.
[ ... ]
> @@ -5122,8 +5123,8 @@ static int query_qp_attr(struct mlx5_ib_dev *dev, struct mlx5_ib_qp *qp,
> if (!outb)
> return -ENOMEM;
>
> - err = mlx5_core_qp_query(dev, &qp->trans_qp.base.mqp, outb, outlen,
> - false);
> + err = mlx5_core_qp_query(qp_to_mdev(qp), &qp->trans_qp.base.mqp, outb,
> + outlen, false);
> if (err)
> goto out;
[Severity: High]
This problem wasn't introduced by this patch, but can query_qp_attr() store
an invalid state in qp->state [2] and cause an out-of-bounds read
in ib_modify_qp_is_ok()?
Just below this call, query_qp_attr() assigns to_ib_qp_state() directly to
the u8 field qp->state:
drivers/infiniband/hw/mlx5/qp.c:query_qp_attr() {
...
qpc = MLX5_ADDR_OF(query_qp_out, outb, qpc);
qp->state = to_ib_qp_state(MLX5_GET(qpc, qpc, state));
...
}
For any hardware QPC state not handled in to_ib_qp_state() (such as
MLX5_QP_STATE_SUSPENDED), to_ib_qp_state() returns -1, which truncates to
255 in qp->state. A subsequent mlx5_ib_modify_qp() call without
IB_QP_CUR_STATE in attr_mask passes cur_state = qp->state (255) to
ib_modify_qp_is_ok(), indexing past the 7x7 qp_state_table[] array.
[1] https://sashiko.dev/#/bug/linux-314256a0-4860-4700-943f-7cc6aed26289
[2] https://sashiko.dev/#/bug/linux-05b234dc-14ed-449f-97cf-d7b4682a7a8d
--
Sashiko AI review · https://sashiko.dev/#/patchset/20261117-mlx5-sd-prep-v1-0-8efd124c7895@nvidia.com?part=4
next prev parent reply other threads:[~2026-11-18 14:04 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-17 14:01 [PATCH rdma-next 00/11] RDMA/mlx5: Preparation for SD support Edward Srouji
2026-11-17 14:01 ` [PATCH rdma-next 01/11] RDMA/mlx5: Enforce dev/mdev naming for device pointers Edward Srouji
2026-11-18 14:04 ` sashiko-bot
2026-11-17 14:01 ` [PATCH rdma-next 02/11] RDMA/mlx5: Rename to_mdev() helpers that return an IB device Edward Srouji
2026-11-18 14:04 ` sashiko-bot
2026-11-17 14:01 ` [PATCH rdma-next 03/11] RDMA/mlx5: Select the PD and SRQ core device through the PD Edward Srouji
2026-11-18 14:04 ` sashiko-bot
2026-11-17 14:01 ` [PATCH rdma-next 04/11] RDMA/mlx5: Select the core device through PD for QP commands Edward Srouji
2026-11-18 14:04 ` sashiko-bot [this message]
2026-11-17 14:01 ` [PATCH rdma-next 05/11] RDMA/mlx5: Cache the MR core device and resolve the MW's via its PD Edward Srouji
2026-11-18 14:04 ` sashiko-bot
2026-11-17 14:01 ` [PATCH rdma-next 06/11] RDMA/mlx5: Select the AH core device through their PD Edward Srouji
2026-11-18 14:04 ` sashiko-bot
2026-11-17 14:01 ` [PATCH rdma-next 07/11] RDMA/mlx5: Select the QP core device through its PD Edward Srouji
2026-11-18 14:04 ` sashiko-bot
2026-11-17 14:01 ` [PATCH rdma-next 08/11] RDMA/mlx5: Resolve core device from ucontext for DEVX resources Edward Srouji
2026-11-18 14:04 ` sashiko-bot
2026-11-17 14:01 ` [PATCH rdma-next 09/11] RDMA/mlx5: Cache the CQ core device and resolve it via cq_to_mdev() Edward Srouji
2026-11-18 14:04 ` sashiko-bot
2026-11-17 14:01 ` [PATCH rdma-next 10/11] RDMA/mlx5: Resolve the bfreg UAR-index core device from the ucontext Edward Srouji
2026-11-18 14:04 ` sashiko-bot
2026-11-17 14:01 ` [PATCH rdma-next 11/11] RDMA/mlx5: Resolve UAR and mmap core devices via " Edward Srouji
2026-11-18 14:04 ` sashiko-bot
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=sashiko-outbox-221133@kernel.org \
--to=sashiko-bot@kernel.org \
--cc=edwards@nvidia.com \
--cc=linux-rdma@vger.kernel.org \
--cc=sashiko-reviews@lists.linux.dev \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox